Add string.find_multiple APIs to pylibcudf (#16920)

Redo at #16824

Contributes to #15162

Authors:
  - Matthew Roeschke (https://github.com/mroeschke)
  - Matthew Murray (https://github.com/Matt711)

Approvers:
  - Vyas Ramasubramani (https://github.com/vyasr)
  - Matthew Murray (https://github.com/Matt711)

URL: #16920

docs/cudf/source/user_guide/api_docs/pylibcudf/strings/find_multiple.rst

@@ -0,0 +1,6 @@
+=============
+find_multiple
+=============
+
+.. automodule:: pylibcudf.strings.find_multiple
+   :members:

docs/cudf/source/user_guide/api_docs/pylibcudf/strings/index.rst

@@ -9,6 +9,7 @@ strings
     contains
     extract
     find
+    find_multiple
     findall
     regex_flags
     regex_program

python/cudf/cudf/_lib/strings/find_multiple.pyx

@@ -1,33 +1,20 @@
 # Copyright (c) 2020-2024, NVIDIA CORPORATION.
 
-from libcpp.memory cimport unique_ptr
-from libcpp.utility cimport move
-
 from cudf.core.buffer import acquire_spill_lock
 
-from pylibcudf.libcudf.column.column cimport column
-from pylibcudf.libcudf.column.column_view cimport column_view
-from pylibcudf.libcudf.strings.find_multiple cimport (
-    find_multiple as cpp_find_multiple,
-)
-
 from cudf._lib.column cimport Column
 
+import pylibcudf as plc
+
 
 @acquire_spill_lock()
 def find_multiple(Column source_strings, Column target_strings):
     """
     Returns a column with character position values where each
     of the `target_strings` are found in each string of `source_strings`.
     """
-    cdef unique_ptr[column] c_result
-    cdef column_view source_view = source_strings.view()
-    cdef column_view target_view = target_strings.view()
-
-    with nogil:
-        c_result = move(cpp_find_multiple(
-            source_view,
-            target_view
-        ))
-
-    return Column.from_unique_ptr(move(c_result))
+    plc_result = plc.strings.find_multiple.find_multiple(
+        source_strings.to_pylibcudf(mode="read"),
+        target_strings.to_pylibcudf(mode="read")
+    )
+    return Column.from_pylibcudf(plc_result)

python/pylibcudf/pylibcudf/libcudf/strings/find_multiple.pxd

@@ -9,5 +9,5 @@ cdef extern from "cudf/strings/find_multiple.hpp" namespace "cudf::strings" \
         nogil:
 
     cdef unique_ptr[column] find_multiple(
-        column_view source_strings,
+        column_view input,
         column_view targets) except +

python/pylibcudf/pylibcudf/strings/CMakeLists.txt

@@ -20,6 +20,7 @@ set(cython_sources
     contains.pyx
     extract.pyx
     find.pyx
+    find_multiple.pyx
     findall.pyx
     regex_flags.pyx
     regex_program.pyx

python/pylibcudf/pylibcudf/strings/__init__.pxd

@@ -9,6 +9,7 @@ from . cimport (
     convert,
     extract,
     find,
+    find_multiple,
     findall,
     regex_flags,
     regex_program,

python/pylibcudf/pylibcudf/strings/__init__.py

@@ -9,6 +9,7 @@
     convert,
     extract,
     find,
+    find_multiple,
     findall,
     regex_flags,
     regex_program,

python/pylibcudf/pylibcudf/strings/find_multiple.pxd

@@ -0,0 +1,6 @@
+# Copyright (c) 2024, NVIDIA CORPORATION.
+
+from pylibcudf.column cimport Column
+
+
+cpdef Column find_multiple(Column input, Column targets)

python/pylibcudf/pylibcudf/strings/find_multiple.pyx

@@ -0,0 +1,39 @@
+# Copyright (c) 2020-2024, NVIDIA CORPORATION.
+
+from libcpp.memory cimport unique_ptr
+from libcpp.utility cimport move
+from pylibcudf.column cimport Column
+from pylibcudf.libcudf.column.column cimport column
+from pylibcudf.libcudf.strings cimport find_multiple as cpp_find_multiple
+
+
+cpdef Column find_multiple(Column input, Column targets):
+    """
+    Returns a lists column with character position values where each
+    of the target strings are found in each string.
+
+    For details, see :cpp:func:`cudf::strings::find_multiple`.
+
+    Parameters
+    ----------
+    input : Column
+        Strings instance for this operation
+    targets : Column
+        Strings to search for in each string
+
+    Returns
+    -------
+    Column
+        Lists column with character position values
+    """
+    cdef unique_ptr[column] c_result
+
+    with nogil:
+        c_result = move(
+            cpp_find_multiple.find_multiple(
+                input.view(),
+                targets.view()
+            )
+        )
+
+    return Column.from_libcudf(move(c_result))

python/pylibcudf/pylibcudf/tests/test_string_find_multiple.py

@@ -0,0 +1,22 @@
+# Copyright (c) 2024, NVIDIA CORPORATION.
+
+import pyarrow as pa
+import pylibcudf as plc
+from utils import assert_column_eq
+
+
+def test_find_multiple():
+    arr = pa.array(["abc", "def"])
+    targets = pa.array(["a", "c", "e"])
+    result = plc.strings.find_multiple.find_multiple(
+        plc.interop.from_arrow(arr),
+        plc.interop.from_arrow(targets),
+    )
+    expected = pa.array(
+        [
+            [elem.find(target) for target in targets.to_pylist()]
+            for elem in arr.to_pylist()
+        ],
+        type=pa.list_(pa.int32()),
+    )
+    assert_column_eq(expected, result)